[MUSIC] In this lesson I'll discuss base lining and assessment. Talk a little bit about what baselining is, why we should use it, explore effective baselining, what it will actually accomplish for you and discuss performance testing and load testing as well. Base line so how do we know that a system is running optimally? This is what we do all the time. So if we want to go out and figure out what the latest performance specs are on a new car, or on a new camera, or anything, that's a standard of how users are interpreting their systems, so cars, or cameras, phones. How do we know that one system is better than another? Well this is baselining, or performance testing. We can apply this to our systems. Baselining takes the shape in a lot of different forms in information technology. We could baseline network performance or storage performance or capacity, or well really all different kinds of things. Frames per second, for example, on gaming, for example, is another way we do baselining, okay? A graphics card should perform at this level. If it's not, then there's something else wrong. This is why we also have resource monitoring on Windows, for example, or the top command using Linux. So we can see how the system is performing and we can do an assessment of what is not performing well Assessments could include a lot of different things. We could assess how a performance is one day, or performance is the next day, or how if we add this much memory, how's it going to effect our systems in this way? Up at our high performance computing data center up in Boulder. We actually one of the engineers, actually looks at their performance and the air flow if we take out a certain, so the servers are built with eight hard drive cages, okay? And if he takes it, takes one out at this area and he takes another one out at this area, another one out of this area, air flow becomes optimal between the servers and we have less cooling that we need to perform. So something like that is to the extreme, but it is baselining the performance of a system. There are a lot of tools out there that we can use to baseline performance. In your reading I've actually included some of those websites that you can go to look for different performance standards or load testing for example. Analyzing and assessing a system for performance is something that we should be doing on a constant basis. For example, when I'm gaming, I always like to see how many frames per second I'm getting based on how I'm over clocking my graphics card, okay? Or the air flow of the system for example. We also see this especially with graphics cards across different brands, across different systems. We have in video coming out with, they just came out with new graphic cards that outperform that even outperform some of the really high end cards last year even. So you may want to go on the internet and look at some benchmarks, some performance tests to see what you should be getting for your systems. For example, think about a car. If you're used to getting 25 miles to the gallon in your car and all of a suddenyou get 20 miles to the gallon, what is going to tell you? It's going to tell you that something is wrong, either wasting too much gas or using too much oxygen or whatever the case is. It's an indication that there's something wrong, I'm receiving poor performance. Okay, load testing should be performed when you're building a system, when you're building a service. How do you know what the baseline of a system can perform at, if you're not load testing? There are also a lot of load testers out there, some are free, some are paid for. For CU, in general, we test the portal speeds. Making sure that how many students can log into the portal at one time to make sure that those systems are always up. On registration day when all 90,000 students for CU register at, I'm sorry, 60,000 students, register at the same time. Can the systems handle that? Can the web servers handle that? Can the database handle that? So if we understand what the baseline of a system is we can understand where our limits are and if they're not performing at that level, what is going on? What change did we make to not meet that level? So in conclusion, baselining and assessment, and constant assessment, is critical to make sure that your systems are running optimally.
